Donor lymphocyte infusions can cure patients with myeloid leukemia in relapse after allogeneic transplant. Few target antigens for these lymphocytes are known, however. As a strategy to search for these antigens, we hypothesize that in myeloid leukemia, peptides derived from normal tissue proteins could be autoantigens for cytotoxic T lymphocytes (CTL) if their overexpression broke immune tolerance or alloantigens for allogeneic CTL if allelic polymorphisms existed between the target cell and the CTL. The long-term goal of this project is to investigate whether HLA-restricted peptides, derived from myeloid selfproteins, can be used to elicit peptide-reactive CTL that preferentially target and kill MDS. The peptides and the peptide-reactive CTL could then be used as vaccines or as adoptive cellular immunotherapy, respectively, to treat patients with MDS. We have shown that normal donors have existent CTL immunity to PR1, an HLA-A2-restricted peptide derived from proteinase 3 (P3). PR1 is an established leukemiaassociated antigen for patients with CML and AML, since active PR1-specific immune responses that kill leukemia correlate with remission. In addition, the results of a recent phase I PR1 peptide vaccine trial show that PR1-specific immunity can reduce the blast count and induce cytogenetic remission in some patients with leukemia.To expand these observations, we.in this Project, will determine whether PR1 peptide vaccine can be used to elicit PR1-CTL immunity in patients with low-risk or minimal disease MDS and whether clinical responses will result from immune response. We will also determine whether additional HLArestricted proteinase 3-derived peptide epitopes can be used to elicit CTL that target myeloid cells from MDS patients and establish an optimal method to expand PR1-specific cytotoxic T lymphocytes (CTL) with high avidity T cell receptors ex vivo.